Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/signal

Pull third pile of kernel_execve() patches from Al Viro:
 "The last bits of infrastructure for kernel_thread() et.al., with
  alpha/arm/x86 use of those.  Plus sanitizing the asm glue and
  do_notify_resume() on alpha, fixing the "disabled irq while running
  task_work stuff" breakage there.

  At that point the rest of kernel_thread/kernel_execve/sys_execve work
  can be done independently for different architectures.  The only
  pending bits that do depend on having all architectures converted are
  restrictred to fs/* and kernel/* - that'll obviously have to wait for
  the next cycle.

  I thought we'd have to wait for all of them done before we start
  eliminating the longjump-style insanity in kernel_execve(), but it
  turned out there's a very simple way to do that without flagday-style
  changes."

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/signal:
  alpha: switch to saner kernel_execve() semantics
  arm: switch to saner kernel_execve() semantics
  x86, um: convert to saner kernel_execve() semantics
  infrastructure for saner ret_from_kernel_thread semantics
  make sure that kernel_thread() callbacks call do_exit() themselves
  make sure that we always have a return path from kernel_execve()
  ppc: eeh_event should just use kthread_run()
  don't bother with kernel_thread/kernel_execve for launching linuxrc
  alpha: get rid of switch_stack argument of do_work_pending()
  alpha: don't bother passing switch_stack separately from regs
  alpha: take SIGPENDING/NOTIFY_RESUME loop into signal.c
  alpha: simplify TIF_NEED_RESCHED handling

2 files changed, 56 insertions, 79 deletions

diff --git a/arch/alpha/kernel/entry.S b/arch/alpha/kernel/entry.S
index 2a359c9ee3cd..a7607832dd4f 100644
--- a/arch/alpha/kernel/entry.S
+++ b/arch/alpha/kernel/entry.S
@@ -311,7 +311,7 @@ entSys:
 
         .align  4
 ret_from_sys_call:
-        cmovne  $26, 0, $19             /* $19 = 0 => non-restartable */
+        cmovne  $26, 0, $18             /* $18 = 0 => non-restartable */
         ldq     $0, SP_OFF($sp)
         and     $0, 8, $0
         beq     $0, ret_to_kernel
@@ -320,8 +320,8 @@ ret_to_user:
                 sampling and the rti.  */
         lda     $16, 7
         call_pal PAL_swpipl
-        ldl     $5, TI_FLAGS($8)
-        and     $5, _TIF_WORK_MASK, $2
+        ldl     $17, TI_FLAGS($8)
+        and     $17, _TIF_WORK_MASK, $2
         bne     $2, work_pending
 restore_all:
         RESTORE_ALL
@@ -341,10 +341,10 @@ $syscall_error:
          * frame to indicate that a negative return value wasn't an
          * error number..
          */
-        ldq     $19, 0($sp)     /* old syscall nr (zero if success) */
-        beq     $19, $ret_success
+        ldq     $18, 0($sp)     /* old syscall nr (zero if success) */
+        beq     $18, $ret_success
 
-        ldq     $20, 72($sp)    /* .. and this a3 */
+        ldq     $19, 72($sp)    /* .. and this a3 */
         subq    $31, $0, $0     /* with error in v0 */
         addq    $31, 1, $1      /* set a3 for errno return */
         stq     $0, 0($sp)
@@ -362,51 +362,35 @@ $ret_success:
  * Do all cleanup when returning from all interrupts and system calls.
  *
  * Arguments:
- *       $5: TI_FLAGS.
  *       $8: current.
- *      $19: The old syscall number, or zero if this is not a return
+ *      $17: TI_FLAGS.
+ *      $18: The old syscall number, or zero if this is not a return
  *           from a syscall that errored and is possibly restartable.
- *      $20: The old a3 value
+ *      $19: The old a3 value
  */
 
         .align  4
         .ent    work_pending
 work_pending:
-        and     $5, _TIF_NEED_RESCHED, $2
-        beq     $2, $work_notifysig
+        and     $17, _TIF_NOTIFY_RESUME | _TIF_SIGPENDING, $2
+        bne     $2, $work_notifysig
 
 $work_resched:
-        subq    $sp, 16, $sp
-        stq     $19, 0($sp)              /* save syscall nr */
-        stq     $20, 8($sp)              /* and error indication (a3) */
+        /*
+         * We can get here only if we returned from syscall without SIGPENDING
+         * or got through work_notifysig already.  Either case means no syscall
+         * restarts for us, so let $18 and $19 burn.
+         */
         jsr     $26, schedule
-        ldq     $19, 0($sp)
-        ldq     $20, 8($sp)
-        addq    $sp, 16, $sp
-        /* Make sure need_resched and sigpending don't change between
-                sampling and the rti.  */
-        lda     $16, 7
-        call_pal PAL_swpipl
-        ldl     $5, TI_FLAGS($8)
-        and     $5, _TIF_WORK_MASK, $2
-        beq     $2, restore_all
-        and     $5, _TIF_NEED_RESCHED, $2
-        bne     $2, $work_resched
+        mov     0, $18
+        br      ret_to_user
 
 $work_notifysig:
         mov     $sp, $16
         bsr     $1, do_switch_stack
-        mov     $sp, $17
-        mov     $5, $18
-        mov     $19, $9         /* save old syscall number */
-        mov     $20, $10        /* save old a3 */
-        and     $5, _TIF_SIGPENDING, $2
-        cmovne  $2, 0, $9       /* we don't want double syscall restarts */
-        jsr     $26, do_notify_resume
-        mov     $9, $19
-        mov     $10, $20
+        jsr     $26, do_work_pending
         bsr     $1, undo_switch_stack
-        br      ret_to_user
+        br      restore_all
 .end work_pending
 
 /*
@@ -454,9 +438,9 @@ $strace_success:
 
         .align  3
 $strace_error:
-        ldq     $19, 0($sp)     /* old syscall nr (zero if success) */
-        beq     $19, $strace_success
-        ldq     $20, 72($sp)    /* .. and this a3 */
+        ldq     $18, 0($sp)     /* old syscall nr (zero if success) */
+        beq     $18, $strace_success
+        ldq     $19, 72($sp)    /* .. and this a3 */
 
         subq    $31, $0, $0     /* with error in v0 */
         addq    $31, 1, $1      /* set a3 for errno return */
@@ -464,11 +448,11 @@ $strace_error:
         stq     $1, 72($sp)     /* a3 for return */
 
         bsr     $1, do_switch_stack
-        mov     $19, $9         /* save old syscall number */
-        mov     $20, $10        /* save old a3 */
+        mov     $18, $9         /* save old syscall number */
+        mov     $19, $10        /* save old a3 */
         jsr     $26, syscall_trace_leave
-        mov     $9, $19
-        mov     $10, $20
+        mov     $9, $18
+        mov     $10, $19
         bsr     $1, undo_switch_stack
 
         mov     $31, $26        /* tell "ret_from_sys_call" we can restart */
@@ -619,24 +603,9 @@ ret_from_kernel_thread:
         mov     $9, $27
         mov     $10, $16
         jsr     $26, ($9)
-        ldgp    $gp, 0($26)
-        mov     $0, $16
-        mov     $31, $26
-        jmp     $31, sys_exit
-.end ret_from_kernel_thread
-
-        .globl  ret_from_kernel_execve
-        .align  4
-        .ent    ret_from_kernel_execve
-ret_from_kernel_execve:
-        mov     $16, $sp
-        /* Avoid the HAE being gratuitously wrong, to avoid restoring it.  */
-        ldq     $2, alpha_mv+HAE_CACHE
-        stq     $2, 152($sp)            /* HAE */
         mov     $31, $19                /* to disable syscall restarts */
         br      $31, ret_to_user
-
-.end    ret_from_kernel_execve
+.end ret_from_kernel_thread
 
 
 /*
diff --git a/arch/alpha/kernel/signal.c b/arch/alpha/kernel/signal.c
index a8c97d42ec8e..32575f85507d 100644
--- a/arch/alpha/kernel/signal.c
+++ b/arch/alpha/kernel/signal.c
@@ -298,8 +298,9 @@ get_sigframe(struct k_sigaction *ka, unsigned long sp, size_t frame_size)
 
 static long
 setup_sigcontext(struct sigcontext __user *sc, struct pt_regs *regs, 
-                 struct switch_stack *sw, unsigned long mask, unsigned long sp)
+                 unsigned long mask, unsigned long sp)
 {
+        struct switch_stack *sw = (struct switch_stack *)regs - 1;
         long i, err = 0;
 
         err |= __put_user(on_sig_stack((unsigned long)sc), &sc->sc_onstack);
@@ -354,7 +355,7 @@ setup_sigcontext(struct sigcontext __user *sc, struct pt_regs *regs,
 
 static int
 setup_frame(int sig, struct k_sigaction *ka, sigset_t *set,
-            struct pt_regs *regs, struct switch_stack * sw)
+            struct pt_regs *regs)
 {
         unsigned long oldsp, r26, err = 0;
         struct sigframe __user *frame;
@@ -364,7 +365,7 @@ setup_frame(int sig, struct k_sigaction *ka, sigset_t *set,
         if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame)))
                 return -EFAULT;
 
-        err |= setup_sigcontext(&frame->sc, regs, sw, set->sig[0], oldsp);
+        err |= setup_sigcontext(&frame->sc, regs, set->sig[0], oldsp);
         if (err)
                 return -EFAULT;
 
@@ -401,7 +402,7 @@ setup_frame(int sig, struct k_sigaction *ka, sigset_t *set,
 
 static int
 setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
-               sigset_t *set, struct pt_regs *regs, struct switch_stack * sw)
+               sigset_t *set, struct pt_regs *regs)
 {
         unsigned long oldsp, r26, err = 0;
         struct rt_sigframe __user *frame;
@@ -420,7 +421,7 @@ setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
         err |= __put_user(current->sas_ss_sp, &frame->uc.uc_stack.ss_sp);
         err |= __put_user(sas_ss_flags(oldsp), &frame->uc.uc_stack.ss_flags);
         err |= __put_user(current->sas_ss_size, &frame->uc.uc_stack.ss_size);
-        err |= setup_sigcontext(&frame->uc.uc_mcontext, regs, sw,
+        err |= setup_sigcontext(&frame->uc.uc_mcontext, regs, 
                                 set->sig[0], oldsp);
         err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set));
         if (err)
@@ -464,15 +465,15 @@ setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
  */
 static inline void
 handle_signal(int sig, struct k_sigaction *ka, siginfo_t *info,
-              struct pt_regs * regs, struct switch_stack *sw)
+              struct pt_regs * regs)
 {
         sigset_t *oldset = sigmask_to_save();
         int ret;
 
         if (ka->sa.sa_flags & SA_SIGINFO)
-                ret = setup_rt_frame(sig, ka, info, oldset, regs, sw);
+                ret = setup_rt_frame(sig, ka, info, oldset, regs);
         else
-                ret = setup_frame(sig, ka, oldset, regs, sw);
+                ret = setup_frame(sig, ka, oldset, regs);
 
         if (ret) {
                 force_sigsegv(sig, current);
@@ -519,8 +520,7 @@ syscall_restart(unsigned long r0, unsigned long r19,
  * all (if we get here from anything but a syscall return, it will be 0)
  */
 static void
-do_signal(struct pt_regs * regs, struct switch_stack * sw,
-          unsigned long r0, unsigned long r19)
+do_signal(struct pt_regs *regs, unsigned long r0, unsigned long r19)
 {
         siginfo_t info;
         int signr;
@@ -537,7 +537,7 @@ do_signal(struct pt_regs * regs, struct switch_stack * sw,
                 /* Whee!  Actually deliver the signal.  */
                 if (r0)
                         syscall_restart(r0, r19, regs, &ka);
-                handle_signal(signr, &ka, &info, regs, sw);
+                handle_signal(signr, &ka, &info, regs);
                 if (single_stepping) 
                         ptrace_set_bpt(current); /* re-set bpt */
                 return;
@@ -568,15 +568,23 @@ do_signal(struct pt_regs * regs, struct switch_stack * sw,
 }
 
 void
-do_notify_resume(struct pt_regs *regs, struct switch_stack *sw,
-                 unsigned long thread_info_flags,
+do_work_pending(struct pt_regs *regs, unsigned long thread_flags,
                  unsigned long r0, unsigned long r19)
 {
-        if (thread_info_flags & _TIF_SIGPENDING)
-                do_signal(regs, sw, r0, r19);
-
-        if (thread_info_flags & _TIF_NOTIFY_RESUME) {
-                clear_thread_flag(TIF_NOTIFY_RESUME);
-                tracehook_notify_resume(regs);
-        }
+        do {
+                if (thread_flags & _TIF_NEED_RESCHED) {
+                        schedule();
+                } else {
+                        local_irq_enable();
+                        if (thread_flags & _TIF_SIGPENDING) {
+                                do_signal(regs, r0, r19);
+                                r0 = 0;
+                        } else {
+                                clear_thread_flag(TIF_NOTIFY_RESUME);
+                                tracehook_notify_resume(regs);
+                        }
+                }
+                local_irq_disable();
+                thread_flags = current_thread_info()->flags;
+        } while (thread_flags & _TIF_WORK_MASK);
 }